Issue #6707 - Make the generated malware domain filter list encode domains as Punycode

sitescripts/subscriptions/test/test_updateMalwareDomainsList.py

Index: sitescripts/subscriptions/test/test_updateMalwareDomainsList.py
diff --git a/sitescripts/subscriptions/test/test_updateMalwareDomainsList.py b/sitescripts/subscriptions/test/test_updateMalwareDomainsList.py
index bd1dd2521aba2c12f550d3dc965c67a78b150386..e63bc5267e461277756e66826b5e2cfeff24a718 100644
--- a/sitescripts/subscriptions/test/test_updateMalwareDomainsList.py
+++ b/sitescripts/subscriptions/test/test_updateMalwareDomainsList.py
@@ -58,7 +58,8 @@ def urlopen(mocker):
         if url.startswith('good'):
             zf_data = io.BytesIO()
             with zipfile.ZipFile(zf_data, 'w') as zf:
-                zf.writestr('justdomains', 'success\n')
+                idna_url = 'xn--xhq02ykwbp4a.cn\n'

[Vasily Kuznetsov, 2018/07/03 16:01:52]

I think it would make sense to have 2 domains in the list just to make sure that
each domain is treated separately by the recoding process. The reason for this
is that if we take something like 'xn--domain1\nxn--domain2\n' and then recode
it without breaking into separate lines, the result is different from breaking
it into separate lines and then processing each line separately (this is what we
want).

The checking in line 77 will need to be a bit more sophisticated to check that
the result is really what we want.

[Tudor Avram, 2018/07/06 11:53:16]

Done.

+                zf.writestr('justdomains', idna_url)
             return io.BytesIO(zf_data.getvalue())
         if url.startswith('bad'):
             raise urllib2.HTTPError('Bad', '42', 'No good', [], None)
@@ -73,7 +74,7 @@ def test_good(md_repo):
     main()
     subprocess.check_call(['hg', 'up'], cwd=md_repo.strpath)
     result = md_repo.join('malwaredomains_full.txt').read()
-    assert 'success' in result
+    assert '.cn-w48dx81cufdl6b' in result
 
 
 def test_bad(md_repo, config):